Lipase-catalyzed synthesis of isoamyl butyrate: Optimization by response surface methodology

Article abstract of DOI:10.1007/s11746-999-0189-x

Immobilized lipase from Mucor miehei (Lipozyme IM-20) was employed in the esterification of butyric acid and isoamyl alcohol to synthesize isoamyl butyrate in n-hexane. Response surface methodology based on five-level, five-variable central composite rotatable design was used to evaluate the effects of important variables - enzyme/substrate (E/S) ratio (5-25 g/mol), acid concentration (0.2-1.0 M), alcohol concentration (0.25-1.25 M), incubation period (12-60 h), and temperature (30-50 °C) - on esterification yield of isoamyl butyrate. In the range of parameters studied, the extent of esterification decreased with temperature, lower E/S ratios, and incubation periods. Excess acid and alcohol concentrations (i.e., acid/alcohol>1.4 or alcohol/acid>1.4) were found to decrease yield probably owing to inhibition of the enzyme by acid or alcohol, the former being more severe. The optimal conditions achieved are as follows: E/S ratio, 17 g/mol; acid concentration, 1.0 M; incubation period, 60 h; alcohol concentration, 1.25 M; and temperature, 30 °C. With these conditions, the predicted value was 1.0 M ester, and the actual experimental value was 0.98 M.

Full text of DOI:10.1007/s11746-999-0189-x

Lipase-Catalyzed Synthesis of Isoamyl Butyrate:
Optimization by Response Surface Methodology
S. Hari Krishnaa, B. Manoharb, S. Divakara, and N.G. Karantha,*
aFermentation Technology and Bioengineering Department and bFood Engineering Department,
Central Food Technological Research Institute, Mysore 570 013, India
ABSTRACT: Immobilized lipase from Mucor miehei (Lipo- back. Considering the high demand and benefits, an opti-
zyme IM-20) was employed in the esterification of butyric acid
and isoamyl alcohol to synthesize isoamyl butyrate in n-hexane.
Response surface methodology based on five-level, five-vari-
able central composite rotatable design was used to evaluate
the effects of important variables—enzyme/substrate (E/S) ratio
(5–25 g/mol), acid concentration (0.2–1.0 M), alcohol concen-
tration (0.25–1.25 M), incubation period (12–60 h), and tem-
perature (30–50°C)—on esterification yield of isoamyl butyrate.
In the range of parameters studied, the extent of esterification
decreased with temperature, lower E/S ratios, and incubation
periods. Excess acid and alcohol concentrations (i.e., acid/alco-
hol > 1.4 or alcohol/acid > 1.4) were found to decrease yield
probably owing to inhibition of the enzyme by acid or alcohol,
the former being more severe. The optimal conditions achieved
are as follows: E/S ratio, 17 g/mol; acid concentration, 1.0 M;
incubation period, 60 h; alcohol concentration, 1.25 M; and
temperature, 30°C. With these conditions, the predicted value
was 1.0 M ester, and the actual experimental value was 0.98 M.
Paper no. J9224 in JAOCS 76, 1483–1488 (December 1999).
mized process with high yields for the economic enzymatic
synthesis of isoamyl butyrate is important.
The goal of the present investigation was to optimize the
process for enzymatic synthesis of isoamyl butyrate using re-
sponse surface methodology employing a five-level, five-vari-
able central composite rotatable design (CCRD). The variables
affecting the esterification that were considered were en-
zyme/substrate (E/S) ratio, acid concentration, incubation pe-
riod, alcohol concentration, and reaction temperature.
MATERIALS AND METHODS
Lipase. Immobilized lipase (triacylglycerol hydrolase, EC
3.1.1.3; Lipozyme IM-20, 25 BIU/g) from Mucor miehei
(presently named Rhizomucor miehei) supported on macro-
porous weak anionic resin beads was kindly provided by
Novo Nordisk (Bagsvaerd, Denmark).
Solvent and substrates. n-Hexane obtained from S.D. Fine
Chemicals (Mumbai, India) was used as the organic solvent.
Isoamyl alcohol was purchased from Fluka Chemie AG (Buchs,
Switzerland). Butyric acid, methanol, and sodium hydroxide
were procured from S.D. Fine Chemicals. All the chemicals
were analytical reagent grade. Soluble impurities, boiling frac-
tions, and excess water were removed by distillation.
KEY WORDS: Central composite rotatable design, enzymatic
synthesis, esterification, isoamyl butyrate, lipase, Lipozyme
IM-20, Mucor miehei, response surface methodology.
Esters of short-chain fatty acids and alcohols are extremely
Esterification. Lipozyme IM-20, which is insoluble in or-
important aroma compounds. For instance, ethyl butyrate and ganic solvents, was employed as a biocatalyst to perform the
isoamyl acetate/butyrate are found, respectively, in the aroma esterification of isoamyl alcohol by butyric acid. Ester syn-
of strawberry and banana. Esters of isoamyl alcohol, espe- thesis was performed in stoppered flasks with a working vol-
cially acetate and butyrate, are valuable, high-demand flavor ume of 10 mL. An appropriate amount of enzyme was added
and fragrance compounds widely used in the food, beverage, into the flask containing freshly prepared solution of isoamyl
cosmetic, and pharmaceutical industries. Currently, most of alcohol and butyric acid dissolved in n-hexane. The reaction
the flavor components are synthesized by chemical methods mixtures were incubated in an orbital shaker (Lab-Line In-
or extracted from plant materials. Recent trends in consumer struments Inc., Melrose Park, IL) at 150 rpm and at specified
preference toward natural products indicate that biocatalysts temperature.
may have an advantage over their chemical counterparts, as
products of biocatalysis may obtain a “natural” label (1).
Determination of esterification yield. Aliquots of the reac-
tion mixture were withdrawn periodically. Samples were as-
Some attempts have been made to evaluate the feasibility sayed both by titrimetry and by gas chromatography (GC).
of producing isoamyl butyrate using lipases from various Samples were titrated against sodium hydroxide to determine
sources (2–4). However, the need for the use of high lipase the residual acid content using phenolphthalein as indicator
and low substrate concentrations has been a significant draw- and methanol as quenching agent. The percentage esterifica-
tion and the moles of acid reacted were calculated from the
values obtained for the blank and test samples. GC analyses
*To whom correspondence should be addressed.
E-mail: ferm@cscftri.ren.nic.in
Copyright © 1999 by AOCS Press
1483
JAOCS, Vol. 76, no. 12 (1999)

1484
S. HARI KRISHNA ET AL.
TABLE 2
TABLE 1
Coded Level Combinations for a Five-Level Five-Variable Central
Composite Rotatable Design (CCRD)
Coded and Actual Levels of Variables Taken for Design of Experiment
Coded level of variable
Test run
no.
Ester
formed (M)
Predicted
ester (M)
Variable
Unit
−2
−1
0
1
2
X1
X2
X3
X4
X5
X1: enzyme/substrate ratioa g/mol
5
10
15
0.6
36
20
0.8
48
25
1.0
60
1
2
3
4
5
6
7
8
9
−1
1
−1
1
−1
1
−1
1
−1
1
−1
1
−1
1
−1
1
−2
2
0
0
0
0
0
0
0
−1
−1
1
−1
−1
−1
−1
1
1
1
1
−1
−1
−1
−1
1
1
1
1
0
0
0
0
−2
2
0
0
0
0
0
0
0
0
0
0
−1
−1
−1
−1
−1
−1
−1
−1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
−2
2
0
0
0
0
0
0
1
−1
−1
1
−1
1
0.161
0.387
0.302
0.275
0.322
0.314
0.124
0.518
0.285
0.319
0.266
0.740
0.213
0.383
0.596
0.426
0.162
0.581
0.185
0.299
0.323
0.553
0.206
0.550
0.557
0.208
0.489
0.476
0.482
0.487
0.479
0.478
0.145
0.397
0.315
0.211
0.287
0.358
0.102
0.527
0.263
0.333
0.275
0.701
0.224
0.475
0.591
0.487
0.204
0.526
0.144
0.326
0.400
0.503
0.245
0.497
0.630
0.274
0.452
0.452
0.452
0.452
0.452
0.452
X2: acid concentration
X3: incubation period
X4: alcohol concentration
X5: temperature
M
h
M
0.2
0.4
24
0.25 0.50 0.75 1.00 1.25
35 40 45 50
12
1
°C 30
−1
−1
1
aEnzyme (g/L)/substrate (mol/L).
1
1
−1
−1
1
1
−1
1
−1
−1
1
0
0
0
0
0
0
0
0
−2
2
0
−1
−1
1
of the samples were performed on a Shimadzu instrument
(model GC 15-A; Shimadzu Corp., Tokyo, Japan) equipped
with a Carbowax 20M column (3 m length, 3.175 mm i.d.)
and a flame-ionization detector. Nitrogen was used as a car-
rier gas with a flow rate of 30 mL/min. Column oven, injec-
tion port, and detector temperatures were maintained at 100,
200, and 250°C, respectively. The percentage esterifications
calculated by GC analysis (which showed product formation)
and by titrimetry (which showed acid consumption) were
found to be in good agreement.
Experimental design. A five-level, five-variable CCRD
was adopted in this study (5). The fractional factorial design
consisted of 16 factorial points, 10 axial points (two axial
points on the axis of each design variable at a distance of 2
from the design center), and 6 center points. The variables
and their levels selected for the study are represented in Table
1. Table 2 shows the actual experiments that were carried out
for developing the model. For creating response surfaces, the
experimental data obtained based on the above design were
fitted to a second-order polynomial equation of the form
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
1
−1
−1
1
1
0
0
−2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
5
4
5
2
[1]
Y = b + b X + b X
+
b X X
ij i j
∑
∑
∑ ∑
0
i
i
ii
i
i=1
i=1
i=1 j=i+1
value (24.5) is very high compared to F10,21 value (3.31). The
coefficient of determination (R2) of the model was 0.921,
which indicates that the model adequately represents the real
relationships among the selected reaction parameters. The
lack-of-fit test (Table 3) also indicates that the model is ade-
quate to represent the experimental data. The average ab-
solute relative deviation is 11.22%. The normal percentage
probability plot (Fig. 1) of the residuals indicates that the er-
rors are normally distributed (R2 = 0.95) and are independent
of each other and that the error variances are homogeneous.
Neglecting the insignificant terms, the final predictive equa-
tion obtained is given as
where Y = ester formed, M; X1 = E/S ratio, g/mol; X2 = acid
concentration, M; X3 = incubation period, h; X4 = alcohol
concentration, M; X5 = temperature, °C; b0 = constant; bii =
quadratic term coefficients; bij = cross term coefficients. The
regression analyses, statistical significances, and response
surfaces were done using Microsoft Excel software (version
5.0; Redmond, WA). Optimization of the reaction parameters
for maximum ester yield were obtained using Microsoft Ex-
cel’s Solver program (version 5.0), which used Newton’s
search method.
RESULTS AND DISCUSSION
Y = 0.452 + 0.080X1 + 0.045X2 + 0.026X3 + 0.063X4
2
2
2
− 0.089X5 − 0.022X1 − 0.054X2 − 0.020X4
+ 0.050X2 X4 − 0.044X2 X5
[2]
The coefficients of the response surface model as given by
Equation 1 were evaluated. Student’s t-test indicated that all
the linear coefficients, all quadratic terms except time and
temperature coefficients, and only the acid–alcohol and
acid–temperature interaction terms were highly significant
(all P < 0.05). The values of the parameters and the analysis
of variance (ANOVA) are presented in Table 3. The ANOVA
indicates that the model is highly significant, as the Fmodel
In the present study, the concentration of acid substrate
was varied and the ester concentration formed after a speci-
fied duration was expressed with respect to conversion of the
acid. While enzyme concentration is an influencing parame-
ter, it is the E/S ratio, in g/mol, that is probably a more im-
JAOCS, Vol. 76, no. 12 (1999)

LIPASE-CATALYZED SYNTHESIS OF ISOAMYL BUTYRATE
1485
TABLE 3
At the lowest concentration of acid (0.2 M) with the low-
est E/S ratio (5 g/mol), esterification was zero (Fig. 2). A
moderate acid concentration (0.6–0.7 M) and high enzyme
concentration (20–25 g/mol) favored maximal esterification
at 0.5 M alcohol concentration in 24 h at 35°C. As the E/S
ratio increased, ester concentration increased at all the acid
concentrations. An increase in acid concentration above 0.7
M resulted in less esterification at any given E/S ratio proba-
bly due to inhibition of the enzyme by the acid beyond 0.7 M
concentration.
Coefficients of the Model and Analysis of Variance (ANOVA)
Regression statistics
Multiple R
R2
0.960
0.921
0.883
0.053
Adjusted R2
Standard error
Observations
32
ANOVA
Degrees
Sum
Mean sum F ratio
P−value
Increase in E/S ratio resulted in increased esterification at
all alcohol concentrations (Fig. 3). As alcohol concentration
increased at any given E/S ratio, ester concentration increased
up to an alcohol concentration of 0.75 M and thereafter de-
creased, probably owing to inhibition of the enzyme by the
higher alcohol concentration. A similar trend was observed at
all enzyme concentrations. Alcohols have been reported to be
competitive inhibitors of lipases, which follow ping-pong bi-
bi kinetic patterns in esterification reactions (8,9). Kinetic
analysis of the present system should throw more light on this
matter.
The effect of varying acid and alcohol concentrations at
constant E/S ratio, temperature, and incubation period (all at
the −1 level, which is equal to 10 g/mol, 35°C, and 24 h, re-
spectively) is shown in Figure 4. At all acid concentrations
from 0.2 to 1.0 M, an increase in alcohol concentration led to
higher yields up to an alcohol concentration of 0.875 M. Fur-
ther increases in alcohol concentration beyond 0.875 M dras-
tically affected the conversions particularly at lower acid con-
centrations, with a maximal drop in esterification in the range
of freedom of squares of squares
Regression
Residual
Total
10
21
31
0.679
0.058
0.737
0.068
0.003
24.493 2.30E−09
Lack of fit
Pure error
16
5
0.053
0.006
0.003
0.001
2.995
N.S.
Coefficientsa
Value
Standard error
t-Stat
P-value
b0
b1
b2
b3
b4
b5
b11
b22
b44
b24
b25
0.452
0.080
0.045
0.026
0.063
−0.089
−0.022
−0.054
−0.020
0.050
0.016
0.011
0.011
0.011
0.011
0.011
0.010
0.010
0.010
0.013
0.013
27.670
7.484
5.23E−18
2.36E−07
3.73E−04
0.026
4.232
2.404
5.859
8.15E−06
4.82E−08
0.035
−8.270
−2.254
−5.611
−2.099
3.762
1.44E−05
0.048
0.001
−0.044
−3.307
0.003
aOnly coefficients significant at P < 0.05 are presented. N.S., not significant.
portant parameter to indicate the behavior, as the requirement of 0.875–1.25 M alcohol. With the increase in alcohol con-
of enzyme is likely to be a function of the substrate (acid) centration (0.875 to 1.25 M), the esterification increased at
concentration. Although enzyme weight percentage (with re-
spect to alcohol and acid) has been used in the literature (6,7),
E/S ratio has been adopted in the present study mainly so as
to neglect the alcohol term and represent the actual quantity
of enzyme present against acyl donor (acid).
FIG. 2. Response surface plot showing the effect of acid concentration,
enzyme/substrate (E/S) ratio, and their mutual interaction on isoamyl
butyrate synthesis. Other variables (alcohol, incubation period, and
FIG. 1. Normal (percentage) probability plot of the residuals.
temperature) are constant at −1 level.
JAOCS, Vol. 76, no. 12 (1999)

1486
S. HARI KRISHNA ET AL.
FIG. 5. Response surface plot showing the effect of alcohol concentra-
tion, temperature, and their mutual interaction on isoamyl butyrate syn-
thesis. Other variables (E/S ratio, acid, and incubation period) are con-
stant at −1 level. For abbreviation see Figure 2.
FIG. 3. Response surface plot showing the effect of alcohol concentra-
tion, E/S ratio, and their mutual interaction on isoamyl butyrate synthe-
sis. Other variables (acid, incubation period, and temperature) are con-
stant at −1 level. For abbreviation see Figure 2.
ent (Figs. 5 and 6). Temperatures above 30°C affected esteri-
higher acid concentrations (>0.6 M). Below an alcohol con- fication negatively. With an increase in temperature from 30
centration of 0.875 M, esterification showed a maximum at to 50°C, esterification decreased at all alcohol concentrations
0.6–0.7 M acid, which decreased slightly at higher acid con- at E/S = 10 g/mol, acid = 0.4 M, and incubation period = 24 h
centrations.
(Fig. 5). At each temperature, esterification increased with in-
The effects of temperature on esterification at varying con- creases in alcohol, up to 0.75–0.875 M, decreasing thereafter
centrations of alcohol or acid were found to be slightly differ- up to 1.25 M alcohol. The decrease in esterification beyond
0.875 M alcohol may be due to enzyme inhibition by alcohol.
FIG. 4. Response surface plot showing the effect of acid concentration, FIG. 6. Response surface plot showing the effect of acid concentration,
alcohol concentration, and their mutual interaction on isoamyl butyrate temperature, and their mutual interaction on isoamyl butyrate synthe-
synthesis. Other variables (E/S ratio, incubation period, and tempera- sis. Other variables (E/S ratio, alcohol, and Incubation period) are con-
ture) are constant at −1 level. For abbreviation see Figure 2.
stant at −1 level. For abbreviation see Figure 2.
JAOCS, Vol. 76, no. 12 (1999)

LIPASE-CATALYZED SYNTHESIS OF ISOAMYL BUTYRATE
1487
Lower temperatures favor esterification at all acid concen-
trations between 0.2 and 1.0 M at E/S = 10 g/mol, alcohol =
0.5 M, and incubation period = 24 h (Fig. 6). However, cer-
tain other features were appreciably different from those ob-
served in Figure 5. At all the temperatures, esterification in-
creased with acid concentration up to a critical value, beyond
which there was a drastic decrease. This critical acid concen-
tration decreased with temperature in the range tested (0.4 M
for 50°C, extent of esterification being 20%; 0.7 M for 30°C,
extent of esterification being 93%). Therefore, lower temper-
atures not only allowed higher acid concentrations to be used
but also resulted in higher conversions. Because further low-
ering of temperature gave very low rates of reaction, 30°C
can be said to be the optimal temperature to use. Also at
higher acid concentrations (>0.7 M) and higher temperatures
(>45°C) zero esterification was observed.
Increases in temperature normally affect the various equi-
librium processes involved in the esterification reaction—
namely, alcohol, acid and ester binding and solubility and
partitioning of the acid between the microaqueous en-
zyme–water–solvent interface and the dissociation equilib-
rium of the acid—which all become more pronounced at
higher temperatures. While the binding equilibria decrease
with the increase in temperatures, acid dissociation and solu-
bility increase with temperature, all resulting in unfavorable
esterification conditions.
Figure 7, depicting the variation of esterification with time,
shows clearly that the period of incubation has only a mar-
ginal effect on the esterification behavior at E/S = 10 g/mol,
acid = 0.4 M, and alcohol = 0.5 M beyond an incubation pe-
riod of 12 h.
FIG. 8. Contour plot showing the ranges of E/S ratios and alcohol con-
centrations to obtain various ester concentrations. Other variables are
constant at their respective levels as follows: acid = +2, incubation pe-
riod = +2, and temperature = −2. For abbreviation see Figure 2.
achieve maximal conversion of the substrate in minimal time
at ambient temperature. For a given temperature and acid con-
centration, the alcohol concentration and E/S ratio required to
attain a known extent of esterification in a given time can be
calculated using Equation 2. Figure 8 shows the contour plot
predicting the extent of esterification for different alcohol
concentrations and E/S ratios. This type of plot is quite useful
experimentally to arrive at economical processing conditions
to obtain the required yield. Table 4 deals with validation of
experimental conditions to obtain required yields indicated
by the contour plot. While several combinations of alcohol
concentrations and E/S ratios can give the same conversion,
from an economic viewpoint, it is desirable to choose the low-
est possible E/S ratio value from Figures 8 and 9 for practical
esterification. For example, 1.0 M ester (representing full es-
terification) could be obtained using about 17 g/mol enzyme
after 60 h, but the same 1.0 M ester can be achieved using
about 24 g/mol enzyme just after 38.3 h (Fig. 9). These opti-
The most efficient, or optimal, condition for the present
system would be to use the lowest amount of enzyme to
TABLE 4
Model Validation Experimentsa
Ester
formed (M)
Predicted
ester (M)
X1
X2
X3
X4
X5
15
10
10
20
20
20
20
20
24
17
0.6
0.4
0.8
0.8
0.8
0.4
0.8
0.8
1.0
1.0
12
24
48
24
48
24
48
24
38
60
0.75
0.5
0.5
0.5
1.0
0.5
0.5
1.0
1.25
1.25
50
35
35
35
35
45
45
45
30
30
0.208
0.243
0.339
0.492
0.768
0.280
0.246
0.420
0.968
0.979
0.222
0.236
0.366
0.476
0.752
0.306
0.262
0.436
1.000
1.000
FIG. 7. Response surface plot showing the effect of temperature, incu-
bation period, and their mutual interaction on isoamyl butyrate synthe-
sis. Other variables (E/S ratio, acid, and alcohol) are constant at −1 level.
For abbreviation see Figure 2.
aAverage absolute relative deviation = 4.697%.
JAOCS, Vol. 76, no. 12 (1999)

1488
S. HARI KRISHNA ET AL.
predicted the esterification reaction between isoamyl alcohol
and butyric acid.
ACKNOWLEDGMENTS
The authors wish to thank the Director, CFTRI, Mysore, for provid-
ing facilities. The author S. Hari Krishna, is thankful to the Council
of Scientific and Industrial Research, India, for the award of Senior
Research Fellowship to carry out research.
REFERENCES
1. Armstrong, D.W., Selective Production of Ethyl Acetate by Can-
dida utilis, in Biogeneration of Aromas, edited by T.H. Parlia-
ment and R. Croteau, ACS Symp. Ser. 317:254–265 (1986).
2. Langrand, G., C. Triantaphylides, and J. Baratti, Lipase Cat-
alyzed Formation of Flavor Esters, Biotechnol. Lett. 10:549–554
(1988).
3. Langrand, G., N. Rondot, C. Triantaphylides, and J. Baratti,
Short-Chain Flavor Esters Synthesis by Microbial Lipases, Ibid.
12:581–586 (1990).
4. Welsh, F.W., R.E. Williams, and K.H. Dawson, Lipase-Mediated
Synthesis of Low Molecular Weight Flavor Esters, J. Food Sci.
55:1679–1682 (1990).
5. Cochran, W.G., and G.M. Cox, in Experimental Designs, 2nd
edn., Wiley, New York, 1992, pp. 335–375.
6. Shieh, C.J., C.C. Akoh, and P.E. Koehler, Four-Factor Response
Surface Optimization of the Enzymatic Modification of Triolein
to Structured Lipids, J. Am. Oil Chem. Soc. 72:619–623 (1995).
7. Shieh, C.J., C.C. Akoh, and L.N. Yee, Optimized Enzymatic Syn-
thesis of Geranyl Butyrate with Lipase AY from Candida rugosa,
Biotechnol. Bioeng. 51:371–374 (1996).
8. Chulalaksananukul, W., J.S. Condoret, and D. Combes, Kinetics
of Geranyl Acetate Synthesis by Lipase Catalyzed Transesterifi-
cation in n-Hexane, Enzyme Microb. Technol. 14:293–298
(1992).
FIG. 9. Contour plot showing the ranges of E/S ratios and incubation
periods to obtain various ester concentrations. Other variables are con-
stant at their respective levels as follows: acid = +2, alcohol = +2, and
temperature = −2. For abbreviation see Figure 2.
mal conditions have also been confirmed experimentally
(Table 4).
Adequacy of the model was also examined at additional
independent conditions that were not employed in this treat-
ment. It was observed that the experimental and predicted val-
ues of ester concentration showed good correspondence
(Table 4). The optimal conditions predicted for synthesizing
1.0 M ester were as follows: E/S ratio = 17 g/mol; acid = 1.0
M; incubation period = 60 h; alcohol = 1.25 M; and tempera-
ture = 30°C. The actual experimental value obtained was 0.98
M, which was in good agreement with the predicted value.
ANOVA also indicated that the generated model adequately
9. Chulalaksananukul, W., J.S. Condoret, and D. Combes, Geranyl
Acetate Synthesis by Lipase Catalyzed Transesterification in Su-
percritical CO2, Ibid. 15:691–698 (1993).
[Received April 29, 1999; accepted August 26, 1999]
JAOCS, Vol. 76, no. 12 (1999)